SignCert-PO mitigates reward hacking in RLHF by down-weighting completions whose advantage signs are not robust to small reward-model perturbations, using a certified preservation radius derived at the policy optimization stage.
F .3.3 Baselines We compare SignCert-PO (Algorithm 1) against the following baselines: •SFT: the SFT policyπ SFT without any RL optimization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Mitigating Reward Hacking in RLHF via Advantage Sign Robustness
SignCert-PO mitigates reward hacking in RLHF by down-weighting completions whose advantage signs are not robust to small reward-model perturbations, using a certified preservation radius derived at the policy optimization stage.